Transplantation Into the Humanbrain: Present Status

J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Journal ofNeurology, Neurosurgery, and Psychiatry. Special Supplement 1989:39-54

Transplantation into the human brain: present status and future possibilities

OLLE LINDVALL From the Department ofNeurology, University Hospital, and Department ofMedical Cell Research, University ofLund, Lund, Sweden

Transplantation of brain tissue into the adult mam- noradrenergic neurons from the locus coeruleus malian CNS is not a new field of research. A century region implanted into the hippocampus suppress ago, the New York Medical Journal contained an seizure development in hyperexcitable, previously article by W Gilman Thompson' entitled "Successful noradrenaline-depleted rats subjected to electrical brain grafting". He exchanged large pieces of neocor- kindling213 (an experimental model ofcomplex partial tical tissue between adult cats and dogs and analysed epilepsy in humans). some ofthe grafts microscopically after survival times However, the transition from the first positive ofup to 7 weeks. Despite the positive title ofhis paper, animal experimental data to trials in patients should the long-term "grafts" probably consisted only of be approached with caution. From the clinical point of neuron-free remnants and developing scar tissue. view, research in neural transplantation has by far Almost 90 years passed before the first animal reached the furthest in Parkinson's disease. Whether Protected by copyright. experimental data appeared clearly indicating the or not intracerebral neural grafting will be successful possible clinical usefulness of the neural grafting in patients with Parkinson's disease will be of great technique. In 1979 it was shown23 that intracerebral importance for the future application ofthis technique grafts ofrat fetal dopamine (DA) neurons were able to in other disease states. Therefore, this review will focus reduce the symptoms of experimental, 6-hydrox- on research in Parkinson's disease and summarises ydopamine (6-OHDA)-induced Parkinsonism in present experience with grafting in patients as well as rats. This was the first example of neurons, being its animal experimental basis. Finally, some issues will implanted into the adult mammalian brain, which be discussed that are ofcritical importance in order to were able to reverse a functional deficit in an animal establish neural transplantation as a real therapeutic model of a human neurological disorder. Early on, of alternative in Parkinson's disease and other course, these findings kindled a hope that it would be neurological disorders. possible to develop a transplantation therapy for patients with Parkinson's disease. Grafting experiments in animals-fetal substantia Since then, experimental data obtained in rats have nigra suggested the clinical application ofthe neural grafting Most basic experimental studies have been carried out

technique also in other disorders. In experimental in rats with unilateral, 6-OHDA-induced lesions ofthe http://jnnp.bmj.com/ Huntington's disease, induced by intrastriatal injec- mesostriatal DA system. Such rats exhibit a hemipark- tions of an excitotoxin (kainic acid or ibotenic acid), insonian-like syndrome with the following main grafted fetal striatum is anatomically integrated into characteristics: the animals have an asymmetric pos- the host brain4 and reduces both motor and cognitive ture and display rotational asymmetry towards the deficits.' In two animal models of dementia (young lesioned side both spontaneously and in response to adult rats with fimbria-fornix transections and cog- amphetamine (which releases DA from the intact nitively impaired aged rats) grafted fetal cholinergic mesostriatal system). Apomorphine, which stimulates neurons from the septum-diagonal band region grow supersensitive DA receptors in the denervated on September 30, 2021 by guest. into the host hippocampus and amelio rate deficits in striatum, causes rotation towards the non-lesioned memory and learning.v" Furthermore, fetal side. The rats also show sensory inattention towards stimuli applied to the side of the body contralateral to Based on the Sir William Macewen Centennial Lecture, Society of British the lesion. Neurological Surgeons, Glasgow, 23 September 1988. Mesencephalic DA-rich grafts implanted directly Address for reprint requests: Olle Lindvall, M.D., Ph.D., Department of into or adjacent to the 6-OHDA denervated striatum Neurology, University Hospital, S-221 85 Lund, Sweden. in the rat can reduce the lesion-induced motor and 1415 Accepted March 1989 sensorimotor asymmetries.23 The degree of 39 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

40 Lindvall behavioural recovery has been found to be related to malities have proved resistant to graft-induced the extent ofthe transplant-derived DA reinnervation recovery. These include deficits in hoarding of the previously denervated striatum."'6 The graft behaviour3" and the aphagia and adipsia seen after tissue, which is taken from the ventral mesencephalon bilateral 6-OHDA lesions39 40 as well as the impairment of 13 to 15 days old rat fetuses, can be implanted into in the use of the contralateral forelimb for discrete adult rats using two main procedures: (1) solid pieces skilled movements after unilateral lesions of the can either be put in the lateral ventricle or into mesostriatal DA system." It has been proposed (see, premade cortical cavities on top of or lateral to the for example, ref41) that the differences in the effects of striatum;23161" (2) a suspension ofcells can be injected grafts on different behavioural measures reflect the stereotaxically directly into the striatum.'" The degree of incorporation of the graft into the host implanted DA neurons grow into the denervated host striatal circuitry. In their ectopic striatal location, striatum forming a terminal pattern similar to the probably only part of the anatomical and functional intrinsic innervation.'9 They also establish synaptic connections characteristic ofnormal mesostriatal DA contacts with host striatal neurons2"22 and receive neurons are re-established by the grafts. The differen- afferent inputs from the host brain.22"24 The grafts are tial sensitivity of different aspects of behaviour to metabolically2" and physiologically3' active; they graft-induced recovery in the 6-OHDA lesioned rat exhibit transmitter synthesis, normal firing patterns, could have important clinical correlates. Nigral grafts and spontaneous DA release. In the host striatum, the may be effective for some but not for all symptoms of graft-derived innervation causes post-synaptic Parkinson's disease in patients. Modifications of the dopaminergic binding sites to return to normal den- transplantation procedures or priming of the grafts sity32 and normalises firing rates and drug sensitivity of pharmacologically38 might be required to overcome host striatal neurons.33 these limitations. Furthermore, DA neurons implanted into specific Successful grafting into the striatum of DA-rich, subregions ofthe denervated striatum compensate for ventral mesencephalic tissue from fetuses has alsoProtected by copyright. specific features ofthe hemiparkinsonian syndrome in been reported in non-human primates with MPTP- 6-OHDA lesioned rats.'34 Amphetamine-induced induced Parkinsonism. Survival of implanted DA rotational asymmetry can be reversed by grafts re- neurons in the caudate nucleus or the putamen has innervating the dorsal caudate-putamen, whereas been demonstrated microscopically in rhesus mon- grafts innervating the ventrolateral part of the keys,'2" African green monkeys4-"' and marmosets.48 striatum ameliorate deficits in sensorimotor attention Biochemical data' have indicated that the grafted but have no effects on rotational asymmetry. The dopaminergic neurons are able to normalise DA functional recovery is only obtained if the DA-rich turnover in previously severely DA depleted areas also mesencephalic grafts are placed near the target area, in non-human primates. Some of the studies have that is, within or directly outside the striatum. reported a long-lasting reduction of the MPTP- Implantation ofDA neurons into the substantia nigra, induced motor abnormalities, including hypokinesia, the actual site of cell body loss after the 6-OHDA rigidity and tremor. lesion, or along the trajectory of the nigrostriatal pathway, has not led to the behavioural effects accom- Grafting experinents in animals-adrenal medulla plished with grafts localised in the striatum.33 This is Soon after the initial reports of functional effects of

most probably due to the inability of the grafted DA fetal DA neurons in experimental Parkinsonism (see http://jnnp.bmj.com/ neurons to extend their axons over a great distance to above), alternative sources of catecholamine-produc- the forebrain target region. ing cells suitable for grafting were also being explored. The functional effects are clearly dependent on the At that time it seemed unclear whether it would ever be survival and continuous presence ofthe graft, since ifit possible, mainly for ethical and immunological is removed or destroyed at any time after transplanta- reasons, to use human fetal tissue for transplantation tion, the deficits immediately reappear.'63663 The graft- in man. Although chromaffin cells from the adult induced recovery from motor asymmetries is specific adrenal medulla have by far attracted the greatest

for DA-rich ventral mesencephalic tissue so that interest, animal experiments have suggested that other on September 30, 2021 by guest. intrastriatal grafts of tissue rich in 5-hydroxytryp- tissues could also be potentially useful for grafting in tamine-producing neurons (from the mesencephalic Parkinson's disease patients, for example, sympathetic raphe region) or of tissue appropriate to the target ganglia,5" 3 carotid body glomus cells52 and cell lines (from the striatum) do not lead to any functional like PCl2.52"-1 compensation.37 Adrenal medulla cells can be implanted into the Although several motor and sensorimotordeficits in lateral ventricle5' " or the DA-depleted striatum ofrats 6-OHDA lesioned animals can be ameliorated by DA- or monkeys.5"' The number of surviving chromaffin rich intrastriatal grafts, other behavioural abnor- cells is, however, low both in rodents52 and particularly J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 41 in monkeys.596263 Normal chromaffin cells secrete data obtained in rodents. Bankiewicz et al63 implanted primarily adrenaline and only low amounts of DA.' adrenal medullary tissue into the DA-denervated In long-term intraventricular adrenal medullary caudate nucleus of unilaterally MPTP-treated rhesus grafts, DA and noradrenaline are the predominant monkeys. Apomorphine-induced rotational asym- catecholamines,65 thus suggesting that transplantation metry was attenuated for 3 months after surgery but causes a shift in the relative amounts of catecho- by 6 months it had returned to control levels. Whether lamines produced by the adrenal medulla. There is deficits in spontaneous behaviours can be reduced by almost no fibre outgrowth from the implanted adrenal medullary grafts in non-human primates is chromaffin cells, which is probably due to inadequate unclear since both improvement of the use of the activity of nerve growth factor (NGF) in the adult affected arm in hemiparkinsonian monkeys63 67 as well striatum. Addition ofNGF to the graft, preferably via as no effect on this parameter4449 have been described continuous infusion through a dialysis fibre for several after surgery. weeks, significantly increases cell survival, causes a transformation ofmany cells towards a more neuronal Grafting experiments in animals-comparison between phenotype and greatly enhances the fibre outgrowth fetal substantia nigra and adrenal medulla into the host striatum.i Survival ofadrenal medullary The available data from basic experimental grafting cells has been demonstrated in the rat for at least 17 studies using either ventral mesencephalic tissue or months after intrastriatal grafting even without addi- adrenal medulla cells clearly favour the strategy with tion of NGF,58 but there appears to be a need for a fetal DA neurons in clinical trials. There seems to be continuous supply of NGF to maintain the graft- three major advantages: First, although both types of derived fibre outgrowth in the host striatum.58 cells can improve deficits in rotational behaviour The functional capacity of adult adrenal medulla apparent after administration of apomorphine or cells implanted either into the striatum or the lateral amphetamine, only nigral cells have been convincingly ventricle has only been demonstrated as an ability to demonstrated to reduce abnormalities in spontaneous Protected by copyright. reduce drug-induced rotational asymmetry in hemi- behaviours. These include spontaneous rotational parkinson rats or monkeys. In unilaterally 6-OHDA- asymmetry, sensory inattention and akinesia in 6- denervated rats, adrenal chromaffin grafts reduce OHDA-lesioned rats and hypokinesia, rigidity and apomorphine-iandpossiblyalsoamphetamine-induced tremor in MPTP-treated monkeys. From the clinical turning response.52 568 The adrenal grafts do not point of view, the capacity of the graft to improve restore contralateral sensorimotor inattention6' and deficits in spontaneous behaviour probably is of no effect on spontaneous rotational behaviour (except critical importance in order to produce relief of during the first hours after implantation) has been symptoms in patients with Parkinson's disease. reported. The reduction of apomorphine-induced Second, in animals with 6-OHDA or MPTP-induced rotational asymmetry is stable up to 3 months after Parkinsonism the fetal nigral grafts show long term grafting but then disappears. However, if NGF is survival and the functional effects are permanent added to the graft through continuous infusion either unless the graft is removed. In contrast, the reduction into the striatum or into the lateral ventricle, the of rotational asymmetry seen after implantation of functional effect (as assessed by the apomorphine adrenal medulla is transient and the chromaffin cells rotation test) is clearly increased58" and more long- show atrophy unless NGF is supplied.

lasting.58 With intrastriatal infusion of NGF in the Third, the fetal DA neurons form appropriate synaptic http://jnnp.bmj.com/ vicinity of the grafts for 4 weeks after implantation, contacts with the host striatum and restore a well- apomorphine-induced rotational behaviour was sig- controlled DA neurotransmission. Adrenal medulla nificantly reduced also at 6 and 12 months after cells, on the other hand, possibly operate via a diffuse, grafting.58 Whether this higher functional potency of non-controlled release of catecholamines (see, adrenal medulla grafts supplied with NGF is due to however, ref 68). This difference could be of critical graft differentiation and neurite outgrowth, as sugges- importance for the usefulness of the grafts in patients ted by Stromberg et al,58 can however, be questioned with Parkinson's disease.

on basis of the recent report by Pezzoli et al.' They on September 30, 2021 by guest. found that NGF plus non-catecholamine-producing Grafting experiments in humans-adrenal medulla tissues (sciatic nerve or fat) were as effective as NGF Despite the encouraging animal data obtained with plus adrenal medulla to reduce apomorphine-induced fetal DA neurons, the first clinical trials with trans- rotational asymmetry in unilaterally 6-OHDA dener- plantation in Parkinson's disease were performed vated rats. using the patient's own adrenal medulla to avoid the The very sparse information reported from ethical and immunological problems linked to the use experiments with adrenal medulla transplantation in of human fetal tissue. Between 1982 and 1985 four non-human primates is in good agreement with the patients with severe Parkinson's disease were sub- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

42 Lindvall jected to adrenal medulla autotransplantation using a stereotaxic procedure69'0 (fig 1). Two of them were grafted in the caudate nucleus and two in the putamen. No adverse effects of the transplantation were observed. There was a minor improvement of motor function lasting for about 2 months but at 6 months no positive effects remained. The long-term follow up of these patients has not indicated any influence of the adrenal medulla autotransplantation on the natural course of the disease. The results from these trials, which did not give anything of long-lasting therapeutic value to the patients, supported the idea that the symptoms of Parkinson's disease can be improved, seemingly with very little risk for the patient, through implantation of " PATIENT'S OWN catecholamine-producing cells in the basal ganglia. It ) ADRENAL GLAND seemed clear, however, that the transplantation procedure had to be improved before any further clinical trials could be carried out. In April 1987 Madrazo and co-workers reported the first successful adrenal medulla autotransplantations Ir in two young patients with Parkinson's disease." 0 0 PIECES OF CHROMAFFIN Instead of a stereotaxic approach Madrazo and TISSUE coworkers used open microsurgical techniques and Protected by copyright. through the cerebral cortex implanted pieces of adrenal medullary tissue into a premade cavity in the head of the caudate nucleus (fig 1). After Madrazo's report was published the number of Parkinson's disease patients subjected to intracerebral grafting increased rapidly (fig 2). Whereas only single cases were operated in 1982 to 1985, about 15 patients received implants in 1986 and about 100 patients in 1987 and 1988. In the majority of patients the adrenal medullary tissue has been implan- STEREOTAXIC OPEN ted into a cavity in the caudate nucleus according to SURGERY MICROSURGERY Madrazo et al,7 but in a few cases successful attempts Fig 1 Schematic representation ofthe two alternative to use stereotaxic surgical techniques have been repor- procedures usedfor grafting ofadrenal medulla cells into the ted.72 Madrazo and co-workers have summarised73 striatum ofpatients with Parkinson's disease. Pieces of their findings from the first 22 patients subjected to chromaffin tissue are either implanted into the caudate adrenal medullary autotransplantation. Four nucleus or putamen using stereotaxic surgical techniques or ofthose http://jnnp.bmj.com/ patients have died due to cerebral venous thrombosis, with open microsurgery into a premade cavity in the head of bronchial aspiration, acute necrotic pancreatitis, and the caudate nucleus. massive pulmonary thromboembolism, respectively.73 Eighteen have been followed clinically for more than without levodopa medication. Madrazo and co-work- one year postoperatively. According to the scoring ers73 have reported that in the 18 surviving cases the system of Madrazo et al73 nine of the patients had a complications of the procedure were transient and poor and nine patients a moderate performance before disappeared within one to two weeks. They included surgery but after transplantation 13 patients had a psychiatric disturbances with hallucinations, stupor, good, four a moderate and one patient a poor delusions, perseveration, respiratory problems, and on September 30, 2021 by guest. performance. From these data it can be concluded that pulmonary thromboembolism. The one exception was at least about 80% of Madrazo's patients have a patient who sustained damage to the fornix during improved after surgery. The improvement has been operation and who had memory impairment. described as a persistent reduction (at least up to 27 About 250 operations using adrenal medullary months) of tremor, rigidity and akinesia both during autotransplantation have been carried out in several on and off periods. The daily doses of levodopa have countries but mainly in Mexico and the United States. been reduced by about 57% and two patients are The scientific reports from these trials are very few and J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 43 TRANSPLANTATION IN PARKINSON S DISEASE

Ln z 70. J 60 o-0~ 50. w 40. m p z 30.

20 Protected by copyright. 10

1982 1983 1984 1985 1986 1987 1988 YEAR Fig 2 Estimated numbers ofParkinson's disease patients subjected to intracerebral grafting ofcatecholamine-producing cells into the striatum. In most cases the implanted tissue was the patient's own adretal medulla but about 50 patients have received grafts ofhumanfetal mesencephalic tissue. the available data mainly derives from the experience dosages of antiparkinsonian drugs could not be in the United States. From the preliminary results of decreased after transplantation. about 90 operations recently reported by different When evaluating the relative value ofthis procedure http://jnnp.bmj.com/ groups in the United States (Third United Parkinson it is ofcourse also important to note that the mortality Foundation Workshop on Brain Implants, Chicago, rate, as well as the serious morbidity rate, as estimated November 1988), it can be stated that only about 40% from the results presented by different groups in the of the patients show some improvement. It has not United States, is as high as 5 to 10%. A majority of been possible to reproduce the dramatic improve- patients show postoperative psychiatric and respi- ments reported by Madrazo and co-workers.7 73 In a ratory disturbances.74 Furthermore, it is a serious detailed analysis7475 of severely disabled Parkinson's drawback with this procedure that the mechanisms of disease patients followed for six months after adrenal improvement are largely unknown. As will be discus- on September 30, 2021 by guest. medulla autotransplantation, performed according to sed below, it seems unlikely that the reduction of the technique ofMadrazo et al,7' a modest reduction of Parkinsonian symptoms is due to catecholamine the duration and severity of "off" periods was release from the implanted cells as originally intended. observed. The mean percentage of "on" time during It is furthermore not clear whether survival ofthe graft the day increased from 48% before transplantation to is necessary. In conclusion, open microsurgical 75% at six months after surgery. Even if the patients autografting of adrenal medulla must be regarded as were more independent during "off>' periods, they an experimental approach and not an established remained prominently affected by their disease. The treatment for Parkinson's disease. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

44 Lindvall Grafting experiments in humans-fetal substantia behaviour (fig 6). The ventral mesencephalon was nigra intact and could be used for implantation into a On the basis of our own experience with adrenal Parkinsonian patient in about 50% of the abortions, medulla autotransplantation in patients6"0 as well as which were carried out by suction (the abortion the more solid data obtained with grafting offetal cells technique used for the fetal ages optimal for DA in experimental Parkinsonism (see above), a pro- neuron survival). Repeated cultures showed that the gramme was initiated in Sweden in 1985 in order to entire procedure could be carried out without bacterial apply the neural grafting technique in patients with contamination. Parkinson's disease. In 1985 the ethical question was In November and December 1987 two 50 year old debated in the Swedish Society of Medicine and this women with Parkinson's disease since 1973 received led to the adoption in 1986 of provisional ethical intrastriatal grafts of human fetal DA neurons (for guidelines for the use of human fetal material for more details see ref 81). Using stereotaxic surgical transplantation purposes. In summary, according to techniques (fig 7) ventral mesencephalic tissue from these guidelines women undergoing abortion must four 8-10 week-old fetuses were implanted as a cell give their consent before tissue material, taken from suspension on the side contralateral to that with the the dead fetuses, can be used for transplantation. The most severe symptoms of Parkinson's disease. The transplantation cannot in any way influence how, suspension was implanted at two sites in the putamen when and why the abortion is carried out. No (tissue from one fetus in each) and at one site in the connection between a particular donor and a recipient caudate nucleus (tissue from two fetuses). The patients is allowed. were put on immunosuppressive treatment at the time In a series of experiments directly related to the oftransplantation using standard regimens ofcyclosp- clinical application ofthe neural grafting technique in orin, azathioprine and prednisolone. The same doses Parkinson's disease, human fetal DA neurons were of antiparkinsonian drugs were given during the six implanted into the DA-denervated striatum of months preceding the transplantation and during theProtected by copyright. immunosuppressed rats."'0 The optimal donor age six months follow-up period. was found to be eight to 10 gestational weeks (fig 3A The results from up to six months postoperatively and B), and ventral mesencephalic tissue from fetuses can be summarised as follows.8' First, there has been of these ages (but not from older ones) survived no marked changes in the percent time of the day the transplantation and had functional effects. The DA patients spend in an "on" phase; Second, small cells reinnervated the entire rat striatum forming a improvements have been noted in clinical tests of the terminal pattern resembling the intrinsic one (fig 4A- motor performance of both patients during off E) and with synaptic contacts with host striatal periods. In patient 1 this slight bilateral improvement neurons (fig 5A, B). The human DA-rich grafts began at about three months after transplantation and spontaneously released DA and reversed deficits in was more marked contralateral to the implanted both drug-induced and spontaneous rotational side. The motor performance of patient 2 was more http://jnnp.bmj.com/ on September 30, 2021 by guest.

Fig 3 (A) The dissected central nervous systemfrom a humanfetus obtained at a suction currettage abortion (9 weeks gestational age). In (B) the ventral mesencephalon (arrow-head) has been dissected out. tv = telencephalic vesicle, m = mesencephalicflexure, p = pontineflexure, c = cervicalflexure. Scale bar = 2 mm. From ref 78. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

...... ~~ ~ ~ ~ ~ ~ ~ .

...... lesion

- z F ..,

;...... _ ;.:-< ...... , ___ 1 liSc=i

:- .--- - }:. - -1. g '. aL!_ e-*i ...... j

*; f j _

._4: ... '...... _-u B jr

ME _---'._.S. _...:.a =fi _

l=:F>-X-:_*:.. .._*, _:: B

D-- EF-.-=.. -.-. .X Protected by copyright. s ._ ....! _- ..... S_ ,T:: -r, - ' .v .o......

.w - C

- ...... E it .!!~~~~~~~~~~~~~~~~~~~~~~~...... / i.: / .'__ t , F: ...... ~~ ~ ~ ~ ~ ~ ~ ~ ~ .. .../.... http://jnnp.bmj.com/ r

;. .

.' :. -. :...

* i

* on September 30, 2021 by guest. f Fig 4 (A-E) Light microscopic pictures ofsections, stainedfor tyrosine hydroxylase (TH) immunocytochemistry to demonstrate preswned DA neurons,from an immunosuppressed rat receiving humanfetal mesencephalic grafts (20 weeks survival after implantation). (A) Overview ofa section through the mesencephalon showing the effects ofthe unilateral 6-OHDA lesion. On the intact side, numerous DA neurons are seen in the substantia nigra (SN) and ventral tegmental area. On the side ofthe lesion virtually no TH-immunoreactivity is visible. (B) Overview ofa section through the grafted caudate- putamen, illustrating the size andposition ofa transplant (t;from a humanfetus, 10-12 weeks gestational age) 20 weeks post- grafting. (C) Higher magnification ofa group ofpresumed DA neurons (arrows) situated within the human mesencephalic graft illustrated in (B). The graft is also seen to have a very denseplexus of TH-positivefibres. Note that the photograph is oriented so that dorsal is to the right and medial is upward. (D) Higher magnification ofa bundle ofcoarse processes, probably dendrites (double arrow heads), which extend into the host striatumfrom the graft. (E) High magnification ofgraft-derived fine-calibre TH-positivefibres (arrowheads) in the host caudate-putamen. Scale bars; (A) I mm; (B) 05 mm; (C) 75 pm; (D) 50 pm; (E) 15 plm. From ref 78. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

46 Lindvall positron emission tomography (PET) performed at 5- *. : : ... 6 months postoperatively, did not show any increase of 6-L-(18F)-fluorodopa uptake in the grafted striatum. The lack of a major clinical improvement in these patients can, in view of the PET data, most likely be ascribed to insufficient number of surviving grafted A-.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. DA neurons. While this could be due to more general problems discussed below (for example, scaling up, immunological rejection, adverse effects on grafted * cells by the continuous antiparkinsonian medication or the disease process itself), it also seems possible that

..... ~~...... - technical factors such as the diameter ofthe implantation instrument and the time interval between the ~~~~~~~~~~~...... abortion and implantation could have been ofimpor- 2F tance. Clinical trials with neural transplantation in Parkin- son's disease are also going on in Mexico,83 England,4 USA, Spain, People's Republic of China, and Cuba. About 50 patients with Parkinson's disease have so far received intrastriatal implants of human fetal DA C-5~~~~~~.... neurons. Both stereotaxic injection of a DA rich cell suspension into striatum (see for example, ref84; fig 7)

and implantation of pieces of ventral mesencephalicProtected by copyright. tissue into a cavity in the caudate nucleus (see ref83; fig 7) are being used. No detailed scientific reports from these studies have yet been published but the preliminary descriptions of the results indicate in some cases much more marked improvements than was observed in the two Swedish patients. The reasons for the discrepancies between the results obtained by different groups are unknown. One major difficulty, at present, is that the procedures used for assessment ofthe patients' symptoms as well as the time the patients have been followed preoperatively Fig 5 (A, B) Photomicrographs showing synaptic differ markedly between various studies. This means specialisations between grafted human DA neurons and host that the reported data are difficult to compare. There striatal neurons at 20 weeks post-grafting. In (A) a have been few attempts to demonstrate survival and TH-positive boutonforms a symmetrical synapse (arrow) ingrowth of grafted DA neurons, for example, with onto a dendritic shaft (d) and in (B) a large PET. Thus, it is not known if the reported TH-immunoreactivefibreforms a symmetrical synaptic improvements are to any extent dependent on restora- http://jnnp.bmj.com/ contact (arrow) with a dendritic spine(s). Scale bars: (A) tion ofDA transmission in the striatum as intended, or and (B) 0-25 gim. From ref 78. due to some other mechanism (see below). It seems unlikely that the early improvements, occurring within variable, but at six months she performed the tests days to weeks, reported by Madrazo et al83 and more rapidly than preoperatively, with no obvious Hitchcock et al,8' are due to a graft-derived reinnerva- side to side difference. However, in neurophysiological tion ofthe striatum. There is in human-to-rat grafting experiments a close correlation between the beginning measurements of simple and complex arm and hand on September 30, 2021 by guest. movements patient 2 showed a significant of fibre ingrowth and synapse formation with host improvement on the side contralateral to transplanta- striatal neurons at about 2-3 months, and the onset of tion but not on the ipsilateral side. Third, the duration functional effects.78 (fig 6). of the response to a single levodopa dose did not change significantly in either patient; fourth, the Possible mechanisms ofaction ofcatecholamine- motor readiness potential, which has been reported to producing cell implants be diminished in patients with Parkinson's disease,82 A major problem when analysing the data reported increased gradually during off in both patients; fifth, from the human trials is that graft survival and graft- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 47 FIBRE INGROWTH AND SYNAPSE 10- FORMATION BEGIN HERE

c E -I c

01- 0 5- : z 0 Protected by copyright.

I I I I I PRE- 10 15 20 TRANS- WEEKS POST-TRANSPLANTATION PLANTATION

Fig 6 Relationship between the time course ofdevelopment offunctional effects and that ofDAfibre ingrowth and synapse formation after implantation ofventral mesencephalic tissuefrom a humanfetus into the DA-depleted striatum of immunosuppressed rats. The amphetamine-induced net rotation asymmetry scores (turns contralateral to the lesion side subtractedfrom turns ipsilateral to the lesion side) are shown both before and at various time points after transplantation for a group ofunilaterally 6-OHDA denervated rats that received DA-rich implantsfrom a 9 week old humanfetus.76 Morphological datafrom ref 78. http://jnnp.bmj.com/ induced reinnervation have not yet been convincingly have shown very little clinical benefit from the opera- demonstrated in any of the operated Parkinson's tion (which agrees with a poor graft survival) and it disease patients. In the few cases that have been can therefore not be excluded that in the more subjected to autopsy (all implanted with adrenal successful cases there is also a higher yield of surviving medullary tissue in the caudate nucleus according to catecholamine-producing cells. Unequivocal demon- Madrazo et al") either no surviving graft8"6 or a very stration ofgraft survival and reinnervation of the host limited number of presumed adrenal medulla cells87 striatum remains, however, one of the major scientific on September 30, 2021 by guest. (chromogranin A-positive but tyrosine-hydroxylase challenges in future clinical trials with transplantation negative) have been demonstrated. Studies performed in Parkinson's disease. with PET in a few patients with grafts of adrenal What are the possible mechanisms of action of medulla' or fetal nigra8" have not detected any signs of catecholamine-rich cell implants underlying the increased catecholaminergic function in the striatum observed improvements in patients with Parkinson's after transplantation (as assessed by the uptake of 18- disease? If other explanations for the clinical changes F-fluorodopa and l-C-nomifensine). It should be can be excluded, such as placebo responses, or mani- pointed out, though, that most of the cases analysed pulations with antiparkinsonian drug treatment, there J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

48 Lindvall

VENTRAL MESENCEPHALON FROM HUMAN FETUS

PIECES OF MESENCEFIHALIC CELL 000 TISSUE SUSPENSION I....' Protected by copyright.

STEREOTAXIC OPEN SURGERY MICROSURGERY Fig 7 Schematic representation ofthe two alternative procedures usedfor grafting ofhumanfetal DA neurons into the striatum ofpatients with Parkinson's disease. The ventral mesencephalon is either prepared as a cell suspension which is implanted stereotaxically into the caudate nucleus and/or putamen, or divided in pieces which are put into a premade cavity in the head ofthe caudate nucleus. are three major possibilities (fig 8A-C): First, the ing adrenal medulla grafting according to Madrazo et grafts could work as biological minipumps and release al" have not demonstrated any significant increases DA (or other catecholamines) diffusely into the even in patients who have exhibited clinical striatum or into the cerebrospinal fluid (fig 8A). This improvement."92 has been reported to occur when rat adrenal medulla Second, the grafted cells may grow into the host cells are implanted into the rat striatum," in which striatum and restore synaptic neurotransmission (fig http://jnnp.bmj.com/ case a leakage oftransmitter into the surrounding host 8B). This mechanism of action seems more likely with tissue can be demonstrated. Such a diffuse cate- implants of fetal neural tissue than after adrenal cholamine release has been proposed to account for medulla autotransplantation (except possibly in a the functional effects observed after adrenal medulla future clinical setting when NGF is administered to autotransplantation, at least in rats. This is supported the graft). There are, however, no PET data available by the finding that chronic infusion of DA into the to support the idea that the improvements reported denervated striatum using a mini-pump is able to after grafting offetal mesencephalon can be attributed reduce apomorphine-induced rotational asymmetry to a graft-derived increase of dopaminergic nerve on September 30, 2021 by guest. to the same degree as intrastriatal adrenal medullary terminal density in the striatum. Furthermore, as grafts.' Indeed, the first human trials were carried out pointed out above, the time course of the with the intention of utilising this mechanism also in improvements reported from the patients operated in the Parkinsonian brain but so far there is little Mexico83 and England' differs markedly from that of evidence that the grafted chromaffin tissue works in ingrowth and synapse formation from human fetal this way in patients except perhaps during a transient DA neurons7778 (at least when xenografted into the rat, period.70 Analyses of catecholamines and their fig 6). metabolites in ventricular cerebrospinal fluid follow- A thirdpossibility is that the graft, at least for a time, J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 49

(A) (C)1-1 DIFFUSE RELEASE TROPHIC ACTION Protected by copyright.

Fig 8 A-C. Schematic illustration ofthree possible mechanisms ofaction ofcatecholamine-producing cell implants in Parkinson's disease. In (A) the two grafts (chromaffin tissue) extendfew processes into the caudate nucleus andputamen ofthe host, and instead act via a difuse release ofcatecholamines into the surrounding striatal tissue. In (B) the grafts (fetal mesencephalic tissue) have grown into the host striatum,formed a new terminal plexus, and reinstated synaptic neurotransmission. In (C) the grafts (chromaffin tissue) show very little outgrowth but have promoted sprouting ofthe host's own DA system leading to an increased density ofDA terminals in the striatum.

exerts a action on the host brain (fig 8C). This trophic http://jnnp.bmj.com/ mechanism, which does not require long-term survival of the graft, could lead to sprouting of the few remaining intrinsic DA neurons or stimulate other recovery mechanisms in the striatum ofthe Parkinson- ian patient. So far, no evidence has been provided that a trophic mechanism underlies the improvements reported in patients either after transplantation of or recent fetal nigral cells, adrenal medulla. However, on September 30, 2021 by guest. studies performed in MPTP-treated mice and monkeys suggest that intrastriatal implantation of fetal nigra or adrenal medulla can indeed promote sprouting of remaining host dopaminergic neurons.49 6393 4 This effect can at least partly be attributed to the (B) parenchymal injury produced by the transplantation technique. In monkeys, adrenal medulla grafts placed SYNAPTIC RELEASE either in a cavity63 4 or stereotaxically9 into the J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

50 Lindvall caudate nucleus induced a presumed sprouting res- is not known. This is, ofcourse, ofgreat importance in ponse of DA fibres to an extent that corresponded to order to decide the number of implantation sites. surgical procedures without any tissue implants. The Before we performed our first grafting experiments in survival of adrenal chromaffin cells in the grafts was patients, we estimated that each human DA neuron minimal.63"' According to Bankiewicz et al49 fetal could grow up to 4a1 to 5 4 mm. Theoretically, about nigral tissue is more effective and gives rise to extensive 40 to 80% ofthe volume ofthe putamen would then be sprouting of presumed dopaminergic neurons in the reached by the growing DA axons implanted along host brain. two injection tracts. The lack ofany major effect could in part be due to an overestimation of the growth Some critical issuesfor successful clinical application capacity ofhuman DA neurons and may indicate that ofthe neural grafting technique a larger number of implantation sites are necessary. Scaling up. Most experimental studies with neural Host environment. A common feature of the transplantation have been carried out in rodents and experimental models of human neurodegenerative only a few in monkeys. In the rat striatum between 100 disorders used in grafting experiments is that the and 200 surviving DA cells are necessary in order to exposure to the causative agent (for example, 6- effect a greater than 50% reduction of amphetamine- OHDA, MPTP) occurs during a short period of time induced rotational behaviour in hemiparkinsonian and has ceased at the time of cell implantation. In rats.95" This number ofcells constitutes only about 1- idiopathic Parkinson's disease the aetiology is un- 2% of the neurons forming the intrinsic mesostriatal known and it may therefore by hypothesised that the DA system in the rat.97 Given the rat data, what is the disease process itselfcould interfere with the long-term critical number of surviving grafted DA neurons survival and outgrowth ofDA neurons implanted into needed in the human brain to obtain a therapeutic the patient's brain. This would most likely not occur if effect? Optimally, 20 000-25 000 DA cells from both the degeneration is due to an intrinsic defect in their sides of the mesencephalon of each human fetus (8 to own DA system, such as a premature or accelerated 10 gestational weeks) survive grafting to the striatum aging process'031 or a lack of scavenger molecules Protected by copyright. ofimmunosuppressed rats." Since it can be estimated98 that can remove endogeneously produced free that the human putamen and caudate nucleus are radicals.'0'07 If, on the other hand, Parkinsonian normally innervated by about 60 000 DA neurons patients lose DA neurons because of a disease process each, grafting of ventral mesencephalic tissue from within the striatum but extrinsic to these neurons, for one fetus into one of these structures might be able to example to a lack of trophic substances, this process restore 30 to 40% of the normal number of cells. The could also elicit a degeneration ofthe grafted neurons. symptoms of Parkinson's disease do not appear until Similarly, if the DA neurons have degenerated due to more than 70% of the DA neurons have degen- an exogeneous insult, of which the causative agent erated9"'°; until this stage is reached, DA transmission were still present, this process could also cause a is believed to be maintained through hyperactivity of gradual death of DA neuron in the graft or could remaining DA neurons and post-synaptic receptor inhibit growth. This would be the case if the patient supersensitivity.'0'102 It therefore seems realistic to continues to be exposed to some environmental toxin postulate that tissue from one or more human fetuses, (see refs 108-110). However, if the degeneration implanted unilaterally into the putamen, caudate resulting from this exposure is a slow process, one nucleus or both, would be sufficient to effect some would expect that the patient should still benefit from symptomatic improvement for a patient with Parkin- DA neuron grafting, at least for a limited time period. http://jnnp.bmj.com/ son's disease. In severely affected Parkinson's disease patients, It should be remembered, though, that the volume drug therapy must be maintained until beneficial of the putamen is 200 times larger in man than in the effects appear, which would be expected at about three rat brain. To what extent this structure will be months postoperatively. However, the grafted reinnervated is not only dependent on the number of neurons may be vulnerable to this treatment, par- surviving grafted DA neurons but also on the number ticularly during the period of nerve fibre formation. It and location ofimplantations and the growth capacity has been suggested that levodopa treatment may of each DA neuron. Grafted rat DA neurons extend accelerate the degeneration of remaining intrinsic on September 30, 2021 by guest. their fibre network about 1 5-2 mm from the graft nigrostriatal DA neurons in Parkinsonian patients border,'9 whereas human DA neurons implanted into (see ref 111). This might occur, for example, through the rat striatum have exhibited a growth distance ofat an autoxidation of levodopa which has been reported least 3 mm," that is until the growing fibres reach the to produce many cytotoxic substances such as free border of the striatum. The human DA neurons thus radicals and quinones."' One cannot rule out the have a higher growth capacity than rat DA neurons possibility that levodopa treatment interferes with the but the maximum extension of their axonal processes survival and growth of the grafted DA neurons; this J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 51 possibility should be tested in animal experiments. disease must be clarified in detail. Above all, it remains Immunosuppression. Although the brain is often con- to be demonstrated that grafted cells can survive sidered to be immunologically privileged (see refs 113 permanently in the human brain affected by various and 114) it is now well established that immunological disease states. Obviously, for further progress in this rejection processes occur also in the CNS.?615 116 Fetal field, it will be vitally important that basic animal neural tissue has been shown to be immunogenic"6117 studies are carried out in parallel to well-designed and can express major histocompatibility complex human trials. (MHC) antigens (the strong transplantation antigens) after the proper inductive stimuli. In fact, in vitro Our own research reviewed in this article was suppor- expression of MHC antigens has been demonstrated ted by the Swedish MRC (14X-8666), Riksbankens after intracerebral grafting. 116 There is evidence ofhost Jubileumsfond, Rut Peterssons Donationsfond and immunisation by fetal neural grafts ofhuman origin in Thorsten and Elsa Segerfalks Stiftelse. I thank Anders a xenograft situation." Immune reactions as well as Bj6rklund, Patrik Brundin, Siv Carlson and Ron fulminant rejections have been observed in animal Mandel for valuable help with the preparation of the recipients of human fetal neural tissue."7 manuscript. However, implantation of allografts of fetal substantia nigra into the brains of MPTP-lesioned non- References human primates has in most cases been performed 1 Thompson WG. Successful brain grafting. NY Med J 1890; without immunosuppression. The available reports 51:701-2. have indicated good graft function up to at least seven 2 Perlow MJ, Freed WJ, Hoffer BJ, Seiger A, Olson L, Wyatt RJ. months postoperatively.47 8 Only one case of regraft- Brain grafts reduce motor abnormalities produced by destruction of nigro-striatal dopamine system. Science 1979;204: ing has been described in monkeys in which a prompt 643-7. loss of graft function was observed on the previously 3 Bj6rklund A, Stenevi U. Reconstruction of the nigrostriatal implanted side."8 It is not known if immunosuppres- dopamine pathway by intracerebral nigral transplants. Brain Protected by copyright. sion is necessary in clinical trials with grafting of Res 1979;177:555-60. at 4 Wictorin K, Isacson 0, Fischer W, Nothias F, Peschanski M, human fetal neuronal tissue, However, it seems Bj6rklund A. Connectivity ofstriatal grafts implanted into the present appropriate to use such treatment in order to ibotenic acid-lesioned striatum. I. Subcortical afferents. optimise the condition for graft survival and to Neuroscience 1988;27:547-62. increase the chance for successful regrafting. If 5 Deckel AW, Robinson RG, Coyle JT, Sanberg PR. Reversal of is long-term locomotor abnormalities in the kainic acid model of immunosuppression is necessary a major problem Huntington's disease by day 18 fetal striatal implants. Eur J that, at present, there are no non-invasive approaches Pharmacol 1983;93:287-8. to monitor rejection ofgrafted fetal tissue in the brain. 6 Deckel AW, Moran TH, Coyle JT, Sanberg PR, Robinson RG. If it precedes any functional effects, rejection would Anatomical predictors ofbehavioural recovery following fetal If positive functional effects striatal transplants. Brain Res 1986;365:249-58. probably pass unnoticed. 7 Isacson 0, Brundin P, Kelly PAT, Gage FH, Bj6rklund A. have developed, rejection would lead to a partial or Functional neuronal replacement by grafted striatal neurons complete disappearance of such graft-induced im- in the ibotenic acid-lesioned striatum. Nature 1984;311: provements. In either situation it would be desirable to 458-60. 8 Isacson 0, Dunnett SB, Bjorklund A. Graft-induced behavioral detect the first signs of a rejection process in the brain recovery in an animal model of Huntington disease. Proc Nati in order to be able to prevent the rejection ofthe graft. Acad Sci USA 1986;83:2728-32.

9 Low WC, Lewis PR, Bunch ST, et al. Function recovery http://jnnp.bmj.com/ Concluding remarks following neural transplantation ofembryonic septal nuclei in adult rats with septo-hippocampal lesions. Nature 1982; Much experimental evidence, derived from several 300:260-2. animal models, suggests that cell transplantation into 10 Gage FH, Bj6rklund A, Stenevi U, Dunnett SB, Kelly PAT. the adult CNS has a great potential to become Intraphippocampal septal grafts ameliorate learning impair- valuable in the treatment of a variety of human ments in aged rats. Science 1984;225:533-6. 1 1 Bj6rklund A, Gage FH. Grafts offetal septal cholinergic neurons neurological disorders. Although grafting of fetal to the hippocampal formation in aged or fimbria-fornix- neurons seems to be the most promising approach at lesioned rats. Ann NY Acad Sci 1987;495:120-36.

present, other sources ofcells, such as adrenal medulla 12 Barry DI, Kikvadze I, Brundin P, Bolwig TG, Bj6rklund A, on September 30, 2021 by guest. or cultured cell lines, may provide alternative Lindvall 0. Grafted noradrenergic neurons suppress seizure development in kindling-induced epilepsy. Proc Nati Acad Sci strategies. In the future, the use of genetically engin- USA 1987;84:8712-5. eered cells for transplantation purposes also seems to 13 Lindvall 0, Barry DI, Kikvadze 1, Brundin P, Bolwig TG, be a realistic possibility."9 However, it must be Bjorklund A. Intracerebral grafting of fetal noradrenergic concluded that at present cell transplantation into the locus coeruleus neurons: evidence for seizure suppression in the kindling model ofepilepsy. Progr Brain Res 1988;78:79-86. human brain is at an early, experimental stage. The 14 Dunnett SB, Bj6rklund A, Stenevi U, Iversen SD. Behavioural mechanisms underlying the reported improvements recovery following transplantation of substantia nigra in rats after transplantation in patients with Parkinson's subjected to 6-OHDA lesions of the nigrostriatal pathway. I. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

52 Lindvall Unilateral lesions. Brain Res 1981;215:147-61. homologous grafts of substantia nigra. Science 1983;222: 15 Brundin P, Bjorklund A. Survival, growth and function of 937-9. dopaminergic neurons grafted to the brain. Progr Brain Res 33 Str6mberg I, Johnson S, Hoffer B, Olson L. Reinnervation of 1987;71:293-308. dopamine-denervated striatum by substantia nigra trans- 16 Bjorklund A, Dunnett SB, Stenevi U, Lewis ME, Iversen SD. plants: Immunohistochemical and electrophysiological Reinnervation of the denervated striatum by substantia nigra correlates. Neuroscience 1985;14:981-90. transplants: functional consequences as revealed by phar- 34 Dunnett SB, Bj6rklund A, Stenevi U, Iversen SD. Grafts of macological and sensorimotor testing. Brain Res embryonic substantia nigra reinnervating the ventrolateral 1980;199:307-33. striatum ameliorate sensorimotor impairments and akinesia in 17 Freed WJ, Perlow MJ, Karoum F, et al. Restoration ofdopamin- rats with 6-OHDA lesions of the nigrostriatal pathway. Brain ergic function by grafting of fetal rat substantia nigra to the Res 1981;229:209-17. caudate nucleus: Long-term behavioural, biochemical and 35 Dunnett SB, Bj6rklund A, Schmidt RH, Stenevi U, Iversen SD. histochemical studies. Ann Neurol 1980;8:510-9. Intracerebral grafting of neuronal cell suspensions. IV. 18 Bj6rklund A, Schmidt RH, Stenevi U. Functional reinnervation Behavioural recovery in rats with unilateral 6-OHDA lesions ofthe neostriatum in the adult rat by use ofintraparenchymal following implantation of nigral cell suspensions in different grafting of dissociated cell suspensions from the substantia forebrain sites. Acta Physiol Scand 1983;Suppl 522:29-37. nigra. Cell Tiss Res 1980;212:3945. 36 Brundin P, Widner H, Nilsson OG, Strecker RE, Bj6rklund A. 19 Bjorklund A, Stenevi U, Schmidt RH, Dunnett SB, Gage FH. Intracerebral xenografts of dopamine neurons: the role of Intracerebral grafting of neuronal cell suspensions. II. immunosuppression and the blood-brain barrier. Exp Brain Survival and growth of nigral cell suspensions implanted in Res 1989;75:195-207. different brain sites. Acta Physiol Scand 1983;Suppl 522:9-18. 37 Dunnett SB, Hernandez TD, Summerfield A, Jones GH, Arbuth- 20 Schmidt RH, Bjorklund A, Stenevi U. Intracerebral grafting of nott G. Graft-derived recovery from 6-OHDA lesions: dissociated CNS tissue suspensions: A new approach for specificity of ventral mesencephalic graft tissues. Exp Brain neuronal transplantation to deep brain sites. Brain Res Res 1988;71:41 1-24. 198 1;218:347-56. 38 Herman JP, Choulli K, Geffard M, Nadaud D, Taghzouti K, Le 21 Freund TF, Bolam JP, Bjorklund A, et al. Efferent synaptic Moal M. Reinnervation ofthe nucleus accumbens and frontal connections of grafted dopaminergic neurons reinnervating cortex of the rat by dopaminergic grafts and effects on the host neostriatum: A tyrosine hydroxylase immunocyto- hoarding behavior. Brain Res 1986;372:210-6. chemical study. JNeurosci 1985;5:603-16. 39 Dunnett SB, Bj6rklund A, Stenevi U, Iversen SD. Behavioural

22 Mahalik TJ, Finger TE, Str6mberg I, Olson L. Substantia nigra recovery following transplantation of substantia nigra in rats Protected by copyright. transplants into denervated striatum ofthe rat: Ultrastructure subjected to 6-OHDA lesions ofthe nigrostriatal pathway. II. of graft and host interconnections. J Comp Neurol Bilateral lesions. Brain Res 1981;229:457-70. 1985;240:60-70. 40 Dunnett SB, Bj6rklund A, Schmidt RH, Stenevi U, Iversen SD. 23 Arbuthnott G, Dunnett SB, McLeod N. Electrophysiological Intracerebral grafting of neuronal cell suspensions. V. properties of single units in dopamine-rich mesencephalic Behavioural recovery in rats with bilateral 6-OHDA lesions transplants in rat brain. Neurosci Lett 1985;57:205-10. following implantation ofnigral cell suspensions. Acta Physiol 24 Doucet G, Murata Y, Brundin P, Bosler 0, Mons N, Geffard M, Scand 1983;Suppl 522:39-47. Ouimet C, Bj6rklund A. Host afferents into intrastriatal 41 Dunnett SB, Whishaw IQ, Rogers DC, Jones GH. Dopamine- transplants of fetal ventral mesencephalon. Submitted to Exp rich grafts ameliorate whole body motor asymmetry and Neurol. sensory neglect but not independent limb use in rats with 6- 25 Schmidt RH, Ingvar M, Lindvall 0, Stenevi U, Bjorklund A. hydroxydopamine lesions. Brain Res 1987;415:63-78. Functional activity ofsubstantia nigra grafts reinnervating the 42 Bakay RAE, Fiandaca MS, Barrow DL, Schiff A, Collins DC. striatum: Neurotransmitter metabolism and ('4C)-2-deoxy-D- Preliminary report on the use offetal tissue transplantation to glucose autoradiography. J Neurochem 1982;38:73748. correct MPTP-induced Parkinson-like syndrome in primates. 26 Schmidt RH, Bjorklund A, Stenevi U, Dunnett SB, Gage FH. Appi Neurophysiol 1985;48:358-61. Intracerebral grafting of neuronal cell suspensions. III. 43 Bakay RAE, Barrow DL, Fiandaca MS, Iuvone PM, Schiff A, Activity ofintrastriatal nigral suspension implants as assessed Collins DC. Biochemical and behavioral correction of MPTP by measurements ofdopamine synthesis and metabolism. Acta Parkinson-like syndrome by fetal cell transplantation. Ann NY Physiol Scand 1983;Suppl 522:19-28. Acad Sci 1987;49S:623-38. 27 Herman Le Moal M. of striatal 44 JP, Choulli K, Activation Bankiewicz K, Plunkett J, Jacobowitz J, et al. Functional and http://jnnp.bmj.com/ dopaminergic grafts by haloperidol. Brain Res 1985;15: histological changes in MPTP hemiparkinsonian monkeys 543-46. with adrenal medullary or fetal mesencephalic brain tissue 28 Rose G, Gerhardt G, Stromberg I, Olson L, Hoffer B. Mon- implants. Abstr. 9th InternationalSymp. on Parkinson 's disease oamine release from dopamine-depleted rat caudate nucleus 1988:43. reinnervated by substantia nigra transplants: An in vivo 45 Redmond DE, Sladek JR Jr, Roth RH, et al. Fetal neuronal electrochemical study. Brain Res 1985;341:92-100. grafts in monkeys given methylphenyltetrahydropyridine. 29 Zetterstr6m T, Brundin P, Gage FH, et al. In vivo measurement Lancet 1986;i:1125-7. of spontaneous release and metabolism of dopamine from 46 Sladek JR Jr, Collier TJ, Haber SN, Roth RH, Redmond DE Jr. intrastriatal nigral grafts using intracerebral dialysis. Brain Survival and growth of fetal catecholamine neurons trans- Res planted into primate Brain Res Bull 1986;17:809-18. 1986;362:344-49. brain. on September 30, 2021 by guest. 30 Strecker RE, Sharp T, Brundin P, Zetterstrom T, Ungerstedt U, 47 Sladek JR Jr, Redmond DE Jr, Collier TJ, et al. Fetal dopamine Bj6rklund A. Autoregulation of dopamine release and neural grafts: extended reversal of methylphenyltetra- metabolism by intrastriatal nigral grafts as revealed by hydropyridine-induced parkinsonism in monkeys. Progr Brain intracerebral dialysis. Neuroscience 1987;22:169-78. Res 1988;78:497-506. 31 Wuerthele SM, Freed WJ, Olson L, et al. Effect of dopamine 48 Fine A, Hunt SP, Oertel WH, et al. Transplantation ofembryonic agonists and antagonists on the electrical activity ofsubstantia marmoset dopaminergic neurons to the corpus striatum of nigra neurons transplanted into the lateral ventricle ofthe rat. marmosets rendered parkinsonian by l-methyl4phenyl- Exp Brain Res 1981;44:1-10. 1 ,2,3,6-tetrahydropyridine. Progr Brain Res 1988;78:479-89. 32 Freed WJ, Ko GN, Niehoff DL, et al. Normalization of 49 Bankiewicz KS, Plunkett RJ, Jacobowitz DM, Kopin IJ, spiroperidol binding in the denervated rat striatum by Cummins A, Oldfield EH. Implant induced catecholaminergic J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

Transplantation into the human brain: present status andfuture possibilities 53 sprouting in MPTP parkinsonian monkeys: behavioral and 69 Backlund E-O, Granberg P-O, Hamberger B, et al. Transplanta- histochemical studies. Soc Neurosci Abstr 1988;5.4. tion of adrenal medullary tissue to striatum in parkinsonism. 50 Itakura T, Kamei I, Nakai K, et al. Autotransplantation of the First clinical trials. J Neurosurg 1985;62:169-73. superior cervical ganglion into the brain. A possible therapy 70 Lindvall 0, Backlund E-O, Farde L, et al. Transplantation in for Parkinson's disease. J Neurosurg 1988;68:955-9. Parkinson's disease: Two cases of adrenal medullary grafts to 51 Pasik P, Martinez JF, Yahr MD, Pasik T, Holstein G, Schanzer the putamen. Ann Neurol 1987;22:457-68. H. Grafting of human sympathetic ganglia into the brain of 71 Madrazo I, Drucker-Colin R, Diaz V, Martinez-Mata J, Torres MPTP-treated monkeys. Soc Neurosci Abstr 1988;5.6. C, Becerril JJ. Open microsurgical autograft of adrenal 52 Bing G, Notter MFD, Hansen JT, Gash DM. Comparison of medulla to the right caudate nucleus in two patients with adrenal medullary, carotid body and PC12 cell grafts in 6- intractable Parkinson's disease. N Engl J Med 1987;316: OHDA lesioned rats. Brain Res Bull 1988;20:399-406. 831-4. 53 Hefti F, HartikkaJ, SchlumpfM. Implantation ofPC12 cells into 72 Jiao S, Zhang W, Cao J, et al. Study of adrenal medullary tissue the corpus striatum of rats with lesions of the dopaminergic transplantation to striatum in parkinsonism. Progr Brain Res nigrostriatal neurons. Brain Res 1985;348:283-8. 1988;78:575-80. 54 Freed WJ, Patel-Vaidya U, Geller HM. Properties of PC12 73 Madrazo I, Drucker-Colin R, Torres C, et al. Long term (more pheochromocytoma cells transplanted to the adult rat brain. than one year) evolution of patients with adrenal medullary Exp Brain Res 1986;63:557-66. autografts to the caudate nucleus for the treatment of Parkin- 55 Jaeger CB. Morphological and immunocytochemical character- son's disease. Proceedings of the XVI CINP Congress, Mun- istics of PC12 cell grafts in rat brain. Ann NY Acad Sci 1987; ich, 15-19 August 1988. Bunney WE. Jr, Hippins H, Laak- 495:334-49. mann G, Schmauss M, eds. 56 Freed WJ, Morihisa JM, Spoor E, et al. Transplanted adrenal 74 Goetz CG, Olanow CW, Koller WC, et al. Multicenter study of chromaffin cells in rat brain reduce lesion-induced rotational autologous adrenal medullary transplantation to the corpus behaviour. Nature 1981;292:351-2. striatum in patients with advanced Parkinson's disease. 57 Freed WJ, Olson L, Ko GN, et al. Intraventricular substantia N Engl J Med 1989;320:337-41. nigra and adrenal medulla grafts: Mechanisms of action and 75 Penn RD, Goetz CG, Tanner CM, et al. The adrenal medullary (3H)spiroperidol autoradiography. In: Bj6rklund A, Stenevi transplant operation for Parkinson's disease: Clinical observa- U, eds. Neural grafting in the mammalian CNS. Amsterdam: tions in five patients. Neurosurgery 1988;22:999-1004. Elsevier, 1985:471-89. 76 Brundin P, Nilsson OG, Strecker RE, Lindvall 0, Astedt B, 58 Str6mberg I, Herrera-Marschitz M, Ungerstedt U, Ebendal T, Bjorklund A. Behavioural effects of human fetal dopamine

Olson L. Chronic implants of chromaffin tissue into the neurons grafted in a rat model of Parkinson's disease. Exp Protected by copyright. dopamine-denervated striatum. Effects of NGF on graft Brain Res 1986;65:235-40. survival, fiber growth and rotational behavior. Exp Brain Res 77 Brundin P, Strecker RE, Widner H, et al. Human fetal dopamine 1985;60:335-49. neurons grafted in a rat model of Parkinson's disease: 59 Morihisa JM, Nakamura RK, Freed WJ, Mishkin M, Wyatt RJ. immunological aspects, spontaneous and drug-induced Adrenal medulla grafts survive and exhibit catecholamine- behaviour, and dopamine release. Exp Brain Res 1988;70: specific fluorescence in the primate brain. Exp Neurol 1984; 192-208. 84:643-53. 78 Clarke DJ, Brundin P, Strecker RE, Nilsson OG, Bj6rklund A, 60 Patel-Vaidya U, Wells MR, Freed WJ. Survival of dissociated Lindvall 0. Human fetal dopamine neurons grafted in a rat adrenal chromaffin cells of rat and monkey transplanted into model of Parkinson's disease: ultrastructural evidence for rat brain. Cell Tissue Res 1985;240:281-5. synapse formation using tyrosine hydroxylase immunocyto- 61 Freed WJ, Cannon-Spoor HE, Krauthamer E. Intrastriatal chemistry. Exp Brain Res 1988;73:1 15-26. adrenal medulla grafts in rats: Long-term survival and 79 Str6mberg I, Bygdeman M, Goldstein M, Seiger A, Olson L. behavioural effects. J Neurosurg 1986;65:664-70. Human fetal substantia nigra grafted to the dopamine- 62 Hansen JT, KordowerJH, Fiandaca MS, Jiao S-S, Notter MFD, denervated striatum of immunosuppressed rats: evidence for Gash DM. Adrenal medullary autografts into the basal functional reinnervation. Neurosci Lett 1986;71:271-6. ganglia of cebus monkeys: graft viability and fine structure. 80 Str6mberg I, Almqvist P, Bygdeman M, et al. Intracerebral Exp Neurol 1988;102:65-75. xenografts of human mesencephalic tissue into athymic rats: 63 Bankiewicz KS, Plunkett RJ, Kopin IJ,Jacobowitz DM, London Immunochemical and in vivo electrochemical studies. Proc WT, Oldfield EH. Transient behavioral recovery in hemipark- Natl AcadSci USA 1988;85:8331-4.

insonian primates after adrenal medullary allografts. Progr 81 Lindvall 0, Rehncrona S, Gustavii B, et al. Fetal dopamine-rich http://jnnp.bmj.com/ Brain Res 1988;78:543-9. mesencephalic grafts in Parkinson's disease. Lancet 1988;ii: 64 Stromberg I, Herrera-Marschitz M, Hultgren L, Ungerstedt U, 14834. Olson L. Adrenal medullary implants in the dopamine-den- 82 Dick JP, Cantello R, Buruma 0, et al. The Bereitschaftsponten- ervated rat striatum. I. Acute catecholamine levels in grafts tial, L-DOPA and Parkinson's disease. Electoencephalogr Clin and host caudate as determined by HPLC-electrochemistry Neurophysiol 1987;66:263-74. and fluorescence histochemical image analysis. Brain Res 83 Madrazo I, Leon V, Torres C, et al. Transplantation of fetal 1984;297:41-5 1. substantia nigra and adrenal medulla to the caudate nucleus in 65 Freed WJ, Karoum F, Spoor HE, Morihisa JM, Olson L, Wyatt two patients with Parkinson's disease. N Engl J Med RJ. Catecholamine content of intracerebral adrenal medulla 1988;318:51.

grafts. Brain Res 1983;269:184-9. 84 Hitchcock ER, Clough C, Hughes R, Kenny B. Embryos and on September 30, 2021 by guest. 66 Pezzoli G, Fahn S, Dwork A, et al. Non-chromaffin tissue plus Parkinson's disease. Lancet 1988;i: 1274. nerve growth factor reduces experimental parkinsonism in 85 Peterson DI, Price L, Small CS. Autopsy findings in a patient who aged rats. Brain Res 1988;459:398-403. had an adrenal-to-brain transplant for Parkinson's disease. 67 Brooks-Eidelberg B, Eidelberg E, Story J, Barret-Tuck F, Boop Neurology 1989;39:235-8. F, Murk S. Adrenal autotransplant in behaviourally trained, 86 Dohan FC, Robertson JT, Feler C, Schweitzer J, Hall C, MPTP-treated monkeys. Soc Neurosci Abstr 1988;157.6. Robertson JH. Autopsy findings in a Parkinson's disease 68 Becker JB, Freed WI. Neurochemical correlates of behavioral patient treated with adrenal medullary to caudate nucleus changes following intraventricular adrenal medfulla grafts: transplant. Soc Neurosci Abstr 1988;7.4. intraventricular microdialysis in freely moving rats. Progr 87 Hurtig HI, Trojanowski JQ, Sladek J, Goldstein M, Joyce J. Brain Res 1988;78:527-33. Adrenal medulla autograft to caudate nucleus for treatment of J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.52.Suppl.39 on 1 June 1989. Downloaded from

54 Lindvall Parkinson's disease. Post mortem analysis of a case. Abstr. 9th system. Acta Physiol Scand 1971;Suppl 367:69-93. International Symposium on Parkinson's disease. 1988:72. 103 Forno LS. The Lewy body in Parkinson's disease. Adv Neurol 88 Perani D, Leenders KL, Tyrrell P, Scarlato G, Frackowiak RSJ. 1986;45:35-43. Adrenal medulla implantation in Parkinson's disease: two 104 Jellinger K. The pathology of parkinsonism. In: Marsden CD, PET studies using 18-F-Dopa and 1 1-C-Nomifensine. Abstr. Fahn S, eds. Movement disorders 2. London: Butterworths, 9th International Symposium on Parkinson's disease. 1988:73. 1987:124-65. 89 Hargraves R, Freed WJ. Chronic intrastriatal dopamine 105 Calne DB, Duvoisin RC, McGeer E. Speculations on the etiology infusions in rats with unilateral lesions ofthe substantia nigra. ofParkinson's disease. Adv Neurol 1983;40:353-60. Life Sci 1987;40:959-66. 106 Kish SJ, Morito C, Hornykiewicz 0. Glutathione peroxidase 90 Burns RS, Allen GS, Tulipan NB. Autologous adrenal medullary activity in Parkinson's disease brain. Neurosci Lett cell transplantation in Parkinson's disease. Abstr. 9th Inter- 1985;58:343-6. national Symposium on Parkinson's disease. 1988:73. 107 Perry TL, Yong VW. Idiopathic Parkinson's disease, progressive 91 Pezzoli G, Silani V, Motti E, Ferrante C. Clinical and bio- supranuclear palsy and glutathione metabolism in the substan- chemical follow-up after adrenal medulla autograft in Parkin- tia nigra of patients. Neurosci Lett 1986;67:269-74. son's disease. Abstr. 9th International Symposium on Parkin- 108 Langston JW, Irwin I, Langston EB, Forno LS. I-methyl4- son's disease. 1988:33. phenylpyridinium ion (MPP+ ): Identification ofa netabolite 92 Franco-Bourland RE, Madrazo 1, Drucker-Colin R, et al. of MPTP, a toxin selective to the substantia nigra. Neurosci Biochemical analyses of lumbar and ventricular CSF from Lett 1984;48:87-92. Parkinsonians before and after adrenomedullary autotrans- 109 Barbeau A, Roy M, Cloutier T, Plasse L, Paris S. Environmental plantation to the caudate nucleus. Soc Neurosci Abstr 1988; and genetic factors in the etiology of Parkinson's disease. Adv 5.10. Neurol 1986;45:299-306. 93 Bohn MC, Cupit F, Marciano F, Gash DM. Adrenal medulla 110 Spencer PS, Nunn PB, Hugon J, et al. Guam amyotrophic lateral grafts enhance recovery of striatal dopaminergic fibers. sclerosis-parkinsonism-dementia linked to a plant excitant Science 1987;237:913-6. neurotoxin. Science 1987;237:517-22. 94 Fiandaca MS, Kordower JH, Hansen JT, Jiao S-S, Gash DM. 111 Melamed E. Initiation of levodopa therapy in parkinsonian Adrenal medullary autografts into the basal ganglia of cebus patients should be delayed until the advanced stages of the monkeys: injury-induced regeneration. Exp Neurol 1988; disease. Arch Neurol 1986;43:402-5. 102:76-91. 112 Graham DG, Tiffany SM, Bell WR Jr, Guthknecht WF. 95 Brundin P, Isacson 0, Bj6rklund A. Monitoring ofcell viability Autotoxidation versus covalent binding of quinones as the

in suspensions of embryonic CNS tissue and its use as a mechanism of toxicity ofdopamine, 6-hydroxydopamine and Protected by copyright. criterion for intracerebral graft survival. Brain Res 1985; related compounds toward C 1300 neuroblastoma cells in 331:251-9. vitro. Mol Pharmacol 1978;14:644-53. 96 Brundin P, Barbin G, Strecker RE, Isacson 0, Prochiantz A, 113 Barker CF, Billingham RE. Immunologically privileged sites. Bj6rklund A. Survival and function of dissociated dopamine Adv Immunol 1977;25:1-54. neurones grafted at different developmental stages or after 114 Widner H, Brundin P. Immunological aspects of grafting in the being cultured in vitro. Dev Brain Res 1988;39:233-43. mammalian central nervous system. A review and speculative 97 Bjorklund A, Lindvall 0. Dopamine-containing systems in the synthesis. Brain Res Rev 1988;13:287-324. CNS. In: Bjorklund A, Hokfelt T, eds. Handbook ofChemical 115 Lund RD, Rao K, Hankin MH, Kunz HW, Gil TJ III. Neuroanatomy, Vol, 2: Classical Transmitters in the CNS. Transplantation of retina and visual cortex to rat brains of Amsterdam: Elsevier, 1984:55-122. different ages. Maturation, connection patterns, and immun- 98 Lindvall 0, Dunnett SB, Brundin P, Bj6rklund A. Transplanta- ological consequences. Ann NY Acad Sci 1987;495:227-40. tion of catecholamine-producing cells to the basal ganglia in 116 Mason DW, Charlton HM, Jones AJ, Lavy CBD, Puklavec M, Parkinson's disease: Experimental and clinical studies. In: Simmonds SJ. The fate ofallogeneic and xenogeneic neuronal Rose FC, ed. Parkinson's disease: clinical and experimental tissue transplanted into the third ventricle of rodents. advances. London: John Libbey & Co Ltd., 1987:189-206. Neuroscience 1986;19:685-94. 99 Bernheimer H, Birkmayer W, Hornykiewicz 0, Jellinger K, 117 Widner H, Brundin P, Bjorklund A, M81ler E. Survival and Seitelberger F. Brain dopamine and the syndromes ofParkin- immunogenicity of dissociated allogeneic fetal neural son and Huntington. J Neurol Sci 1973;20:415-55. dopamine-rich grafts implanted into the brains of adult mice. 100 Riederer P, Wuketich S. Time course of nigrostriatal degenera- Exp Brain Res 1989 (in press). tion in Parkinson's disease. JNeural Transm 1976;38:277-301. 118 Freed CR, RichardsJB, HuttCJ, Kriek EH, Reite ML. Rejection 101 Agid Y, Javoy F, Glowinski J. Hyperactivity of remaining of fetal substantia nigra allografts in monkeys with MPTP- http://jnnp.bmj.com/ dopaminergic neurons after partial destruction of the nigro- induced Parkinson's syndrome. Soc Neurosci Abstr 1988;7.7. striatal dopaminergic system in the rat. Nature 1973;245: 119 Gage FH, Wolff JA, Rosenberg MB, et al. Implantation of 150-1. genetically engineered cells to the brain. Progr Brain Res 1988; 102 Ungerstedt U. Postsynaptic supersensitivity after 6-hydroxy- 78:651-8. dopamine induced degeneration ofthe nigrostriatal dopamine on September 30, 2021 by guest.